This LoRA was trained to give images a creepy, kitschy vibe, like the name suggests.
Sometimes, you'll get much more than just a vibe.
Description
A number of bizarre deformations are baked into this version.
FAQ
Comments (5)
Hello,
Please, which model do you use?
Which model do I use for what?
@ACLander Well, the model you used to make the images you show as examples. I ask because I haven't been able to imitate their aestheticism.
@bcauldron I trained the LoRA on Pony V6 using Civitai's trainer, and these are the epoch sample images, so V6 would probably get you the closest results (see the metadata on the image pages). But the fact is the images you posted using other checkpoints are equally good.
@bcauldron Also, sometimes longer prompts can outweigh a LoRA at normal strength, so you might want to try upping the weight of the LoRA.